The functionality and user acceptance of prosthetic limbs and myoelectric prostheses in particular can be improved by providing sensory feedback. This project aims to develop a sensory system which prosthetists can incorporate into (or retrofit to) presently available myoelectric arms. The system will be non-invasive and is based on using an array of vibrotactile actuators to input coded information to the skin of the residual limb. Three categories of sensory information will be targeted for powered arms: grasp force, object slippage and either wrist rotation (i.e. pronation/supination) or span of finger opening. Prior work in developing vibrotactile strategies for sensory substitution has shown it is a promising approach. The implementation of that research into the marketplace, however, is being held up by a lack of suitable vibration actuators. The Specific Aim of this PHASE 1 project is to develop vibration actuators specifically for use with artificial limbs. The designed actuators will have minimal bulk and mass; consume only modest power; be inexpensive to fabricate; and require no maintenance. This project will develop novel solenoid type actuators, because optimal vibrotactile feedback codes require varying the vibration frequency and amplitude independently, and this can not be achieved using typical cell phone and pager rotatory motor based vibrators where the frequency and amplitude co-vary. The PHASE 1 work will develop prototypes of the novel actuators and test them for their suitability as input devices for vibrotactile sensory feedback codes for use with upper and lower extremity prosthesis users. In PHASE 2, the sensory feedback system will be extensively tested with upper extremity amputee subjects to optimize the coding strategy; validate that the system adds functionality to the prostheses; and demonstrate the reliability of the system components. Additionally, during PHASE 2, we will expand the user base to include lower extremity prosthesis users, where foot-floor contact and knee position information will be targeted.